Nov 8, 2012

Golson's Progress: Part 3


Part 3 of this installment series takes a look at the three most recent games:  BYU, Oklahoma, and Pitt.  Sadly for the comparison (thankfully for the amount of plays I had to review), a concussion kept Golson from playing at all in the BYU game.  The following 2 games were truly something to remember.  First, an epic showdown in Norman, OK in a game that every pundit in the nation had pre-determined the Irish would lose.  Then, a non-intentionally epic game versus Pitt that caused all of us watching to lose a few years of our lives.  Before I get to the numbers, one thing, while my readership is quite minimal, if any of you who read this are the incredibly kind and dedicated people who post condensed games onto Youtube, please get back at it.  Since no condensed video was available, I had to scroll back-and-forth through online full games which just sucks.  Now that my complaining is out of the way, let’s add on Segment 3’s games to the normal statistical breakdown I’ve used to start the last couple of pieces:

Category:
Games 1-3:
Games 4-6:
Games 7-9:
Attempts:
81
54
67
Completions:
47
32
36
Completion %:
58%
59%
54%
Yards:
611
357
404
TD’s:
3
1
2
INTS’s:
1
2
1
Yards per completion:
13
11.16
11.22
Yards per attempt:
7.54
6.61
6.03

For the most part, everything stayed more or less in line with Segment 2, and slightly down from Segment 1.  On yesterday’s post, commenter “Dan” suggested that the increased difficulty of the defenses played on a game-to-game basis might explain this.  My gut tells me he’s correct.   I think that I may do one “wrap up” piece either Thursday or Friday to explore that more thoroughly by considering Golson’s game-to-game numbers in view of the relative strength of the defense he was up against.

On the surface, one aspect that sticks out to me about the numbers above is the slight drop off in completion percentage.  In theory, a quarterback progressing will get better in this regard as he sees more live action.  Most Notre Dame fans would consider Oklahoma to be Golson’s best all-around game to this point in the season.  Versus a very good secondary, Golson avoided any major mistakes even if he didn’t record a touchdown through the air. He was also 13/25 in that game.  (aside:  I will address his improvements running later.)

In part one of this series, I was critical of Golson’s reluctance to throw the ball away.  His tendency to rely upon his scrambling ability and to attempt to make something out of nothing bothered me.  The area he progressed in the most in games 7-9 was doing just that….throwing the ball away.   I wanted to at least  consider what happens when we take out the throwaways from attempts with respect to completion percentage.  It’s not a good thing that he’s having to throw the ball away, and Golson’s reluctance to make quick decisions certainly plays into the number of times he has to decide between forcing a pass and conceding the play, but for a young quarterback, learning that it’s okay to throw it away is definitely a skill to appreciate:

Category:
Games 1-3:
Games 4-6:
Games 7-9:
Attempts:
81
54
67
Throwaways:
4
3
6
Effective Attempts:
77
51
61
Completions:
47
32
36
Completion %:
61%
63%
59%

Does Golson make up all of the ground on completion percentage once we take out the throwaways?  No.  However, he does make up a little ground, and should give us some (moderate) hope of improvement in decision-making going forward.  I was happy to see that the number of forced throws was in fact decreasing over time.

Okay, let’s get to the throwing by down breakdown:

Down:
Attempts:
Completions:
Comp. %:
TD’s:
INTs:
1
30
15
50%
2
0
2
18
14
78%
0
1
3
19
7
37%
0
0
4
0**
0
---
0
0
·         ** Golson did attempt a throw on 4th down:  The now infamous pass interference against Eifert during the Pitt game.  Of course, it was a P.I., and so no official attempt is recorded.

Rather than explaining this in paragraphs, let’s do the side-by-side comparison in a couple of different ways.  First, by completion percentage by down for each of the 3 segments:

Down:
Games 1-3
Games 4-6
Games 7-9
1
51%
64%
50%
2
68%
62%
78%
3
56%
50%
37%

I’ve highlighted the top performer for each of the 3-downs (excluding 4th down given that no official attempts have been registered).  I don’t think there’s too much of a takeaway.  Random fluctuation probably matters just as much as anything else.  The only unsettling aspect is Golson’s 20 point drop in completion percentage on third-downs from Segment 1 to Segment 3.  It’s also unnerving that that’s the only category where a downward pattern developed.  More on that shortly.  I also wanted to consider the percentage of attempts that came on each down for each segment:

Down:
Games 1-3**
Games 4-6
Games 7-9
1
46%
46%
45%
2
35%
24%
27%
3
20%
30%
28%
·        ** Rounding errors apply.

One observation about first downs, and then I'm moving to the more important second and third down situations:  It’s interesting how consistent the percentage of Golson’s attempts have occurred on first down across the different game periods.  That said, I think this is just a matter of coincidence.  Even from a schematic standpoint, comparing Segment One and Two’s games which both involved blow outs where Notre Dame’s offense shifted strongly towards the run in the second half to Segment Three where the Oklahoma game was in doubt until the fourth quarter and the Pitt game required a near miraculous come from behind victory is difficult to say the least, and more importantly, somewhat meaningless.

Golson was not asked to pass as much in third down situations in Segment One which makes sense for a young starter.  The standout game was the Dublin opener versus Navy in which Golson had only one throwing attempt on third down (The team only had 10 third downs the entire game).  Though, even in the Purdue game (26%) and Michigan State game (21%) the team did not ask him to make a lot of third down throws compared to what we'd see going forward.

Perhaps if we go to the “money down” down and distance chart for Segment Three we’ll get something better:

Down:
Attempts:
Completions:
Comp. Percent.
TD’s:
INT:
2nd 5+
11
8
73%
0
1
2nd <5
7
6
86%
0
0
3rd 5+
12
5
42%
0
0
3rd <5
7
2
29%
0
0

I’ve highlighted second and short for a very particular reason:  In Part 2 of this series, I said it was absolutely imperative that Golson get more opportunities to pass in that situation.  He finally did.  In case you’re wondering, yes, the incredibly important play action 50 yard strike to Chris Brown in the Oklahoma game occurred on 2nd and short.  The regret?  Golson’s other 5 completions on second and short amassed 27 yards combined.  Sadly, his increased opportunities were generally used for short completions to move the chains, which he mostly did well.  Given his third down numbers in these games, maybe Kelly was on to something, but I’d still like to see 2nd and short be used more aggressively going forward.

I also wanted to touch on Golson's drop off in third down success rate.  I'd thought that perhaps he just didn't do as well on third and long plays, but that wasn't true.  His completion numbers on third and long remained fairly similar, but his decline in performance on third and short was awful. Through the first 6 games of the season Golson was a combined 5/8 on third and less than 5 which comes out to a 63% completion rate.  In the last two games, he's been in third and short situations more frequently, and has not faired as well.

So why  has Golson been in more second and short and third and short situations?  The first chart above demonstrates that Golson continued to see a decline in yards per completion and yards per attempt, and I think when we consider the travel distance of passes through the air, the reason reveals itself:

Length:
Attempts:
Completions:
Comp. Percent.
TD’s
INT:
< 10
39
27
69%
2
1
10+
22
9
41%
0
0

Let’s review the comparisons by Segment.  First, completion percentage for each:

Length:
Games 1-3:
Games 4-6:
Games 7-9:
< 10
68%
73%
69%
10+
46%
52%
41%

And then, percentage of pass attempts travelling each distance:

Length:
Games 1-3:
Games 4-6:
Games 7-9:
< 10
66%
51%
64%
10+
34%
49%
36%

Golson’s pattern reverted (not necessarily regressed) to Segment 1 in terms of breakdown between passes travelling less or more than 10 yards through the air.  However, he was worse in terms of completing plays down the field than at any other time.  Part of this is explained by the Pitt game where in the scramble to come back, the team took more shots for the big play (very rarely successfully).  I liked to see, particularly in the Pitt game, that the use of the bubble screen and slip screen to wide receivers came back into use more prevalently, but the team was not very effective at breaking away (aside:  with the exception of the great slip screen to T.J. Jones in the Pitt game that went for an 11 yard touchdown at as pivotal a moment as can be imagined).  Golson’s longest completion on a pass travelling less than 10 yards through the air was just 14 yards.  17 of Golson's 39 attempts of short passess occurred on first down, and only 4 gained 10 or more yards.  While he wasn’t much better in the previous periods, the lack of short-pass, big-play yardage bothers me.  It’d be nice to see our passing game execute on a short pass and make a much larger play to take some pressure off of EG. 

Additionally, the passing game continued to be generally ineffective even when Golson completed the ball deeper.  Yes, he converted on two remarkably important plays:  The 50 yard strike to Brown previously mentioned, and the 45 yard jump ball to Daniels in the Pitt game, but Golson had only one other play on passes of ten yards or greater through the air that netted 20 yards.  Amassing just 3 plays of 20+ yards in the passing game was his fewest for any of the three Segments.

Finally, before I get to the general observations section, I want to make sure I cover targets by position:

Target:
Attempts:
Completions:
Comp. Percent.
TD’s:
INT:
WR:
37
24
65%
1
0
RB:
7
4
57%
1
0
TE:
17
8
47%
0
1

Not much changed on this front.  Receivers still led the way in targets, and all three categories remained fairly static.  Running backs were targeted slightly less, but not in any meaningful manner.  We’re basically talking about a target or two between all of the subsets to put the numbers directly in line with previous Segments. 

Like I said, I’m going to do a wrap-up where I’ll take some additional space to think things over and provide a few more incites into what I’ve watched.  The tight end completion percentage number continued to fluctuate.  The team’s goal of getting Eifet touches sometimes plagues their decision-making.  Golson’s one interception from this Segment came on a pass to a tight end…only it wasn’t Eifert.  On the fateful play that most of us thought would derail Notre Dame’s season, Golson overlooked an open Eifert over the middle and instead released a horribly underthrown ball to Troy Niklas in the corner of the end zone which was picked. 

I’ll do a larger general breakdown on Golson for Friday, so I’ll keep my general observations very short:

1.  Golson appears much more comfortable in the empty backfield set when he sets up to pass.  He scrambles less and seems less anxious.  My general impression in this respect is that he is very uncomfortable either dropping back or having to do play action fakes.  In general, his drop back footwork is lazy.  The empty backfield, shotgun formation alleviates all of these issues.  He doesn’t have to drop back and simultaneously get himself into a good throwing position, and he doesn’t have to worry about the play fake.

2.  Golson can throw the left side of the field.  His biggest issue in this respect is that unless the play is a quick strike designed to go to the left side, he just generally ignores what’s going on over there.

3.  I’ll have more to say about this in the wrap-up, but I fully appreciate that I’ve ignored Golson’s running ability to this point in my analysis.  Partly, this is intentional.  I wanted to focus on his passing.  Many of Golson’s runs through the first set of games were the result of scrambles on broken plays.  In both the OU and Pitt games, the number of designed QB runs increased, and it played to perhaps Golson’s greatest strength.

Can’t wait to do the wrap-up as that’s what all of this research was intended to generate anyways.  If there’s anything specific you’d like to see covered, hit me up at JLDthoughts@gmail.com.

Nov 7, 2012

Golson's Progress: Part 2


Part 2 focuses on games 4-6 of Golson’s year:  Michigan, Miami, and Stanford.  These were troubling times for our QB.  The Michigan game featured just 8 passes (and 2 INT’s) before Tommy Rees relieved him.  Stanford’s hard hitting defense managed to knock him out in the 4th quarter with a concussion that would have a lasting impact.  Given both of these circumstances, the number of pass attempts upon which we can evaluate his performance decreased considerably.  Games 1-3, Golson attempted 81 passes.  That number fell to 54 attempts in games 4-6.  Since the goal is to evaluate Golson’s progress, there will be a lot of side-by-side charts intended to reflect the differences between games 1-3 and 4-6.  Let’s start off by looking at the superficial numbers broken up by these games:

Category:
Games 1-3:
Games 4-6:
Attempts:
81
54
Completions:
47
32
Completion %:
58%
59%
Yards:
611
357
TD’s:
3
1
INTS’s:
1
2
Yards per completion:
13
11.16
Yards per attempt:
7.54
6.61

Completion percentage remained virtually unchanged.  Yet, Golson was less effective in every other category.  He threw fewer touchdowns, more interceptions, his yards per completion dropped, and his yards per attempt dropped.  Even accepting that circumstance and injury limited his attempts, he was not as good on a per attempt basis in any measureable category.  In games 1-3, Golson had 11 passing plays that amassed 20 net yards or more.  In games 4-6, he had just 5.  As a percentage of attempts, that comes out to:  Games 1-3:  13.58%; Games 4-6:  9.26%.  Golson’s longest passing play in games 4-6 was 24 yards.  In games 1-3, he had 7 passes that surpassed that.  We’ll likely get a chance to fill this in better through the remainder of the article, but the net result:  Golson was equally effective at completing passes, but less likely to get a big play.

The next aspect of Golson’s play that I looked at previously was his performance by down.  Here’s the comparison in those categories:

Down:
Attempts:
Completions:
Comp. %:
TD’s:
INTs:
1
25
16
64%
0
1
2
13
8
62%
0
1
3
16
8
50%
1
0
4
0
0
0
0
0

Golson actually got better on first down completion percentage, but not a lot else.  He declined in completion efficiency both respect to second and third downs.  There was only a slight decrease in the percentage of attempts on first down.  In the first third, 47% of Golson’s attempts occurred on first down.  In the second third?  46%.  CBK seems to have given Golson more leeway on third down to throw the ball.  The number of attempts was exactly the same despite the decreased number of overall attempts.  During the first third, 88% of Golson’s third down passing attempts were on 3rd and 5 or more.  By contrast, in the second third, the rate of third and longs (defined as 3rd and 5 or greater) was just 63%.  Kelly, for better or worse, gave more leeway to Golson on the most important down in order to convert pivotal attempts to keep the drive going.  Let’s look at the breakdown:

Down:
Attempts:
Completions:
Comp. Percent.
TD’s:
INT:
2nd 5+
12
7
58%
0
1
2nd <5
1
1
100%
0
0
3rd 5+
10
4
40%
1
0
3rd <5
6
4
67%
0
0

Just as a matter of comparison, let’s consider the percentage of Golson’s total pass attempts during each segment in relation to these particular down and distance situations.

Down:
% 1-3
% 4-6
2nd 5+
28%
22%
2nd <5
6%
2%
3rd 5+
17%
18%
3rd <5
2%
11%

The small sample-size makes these comparisons somewhat artificial, but the thing that sticks out to me was the increased number of throws Golson was entrusted with on third and short.  Yes, in absolute numbers, 6 compared to 2 is not that significant, but it’s at least something worth noting.  What we’re tracking is Golson’s throws and when they occur. 

While I don’t have the full number of opportunities available, I wish the Irish offense was giving Golson more opportunities on second and less than 5.  6 attempts through the first six games is just not good enough.  One thing that all of us as fans would like to see is Golson get a few more explosive deep plays.  Second and short is the best opportunity to do so.  Be it a lack of opportunities or conservative rush-heavy play calling, averaging just 1 attempt per game in the second and short situation through 6 games does Golson no favors. 

Another interesting observation:  66% (2/3) of Golson’s interceptions have occurred on throws on second and long.  Perhaps this is just coincidence, but second and long is the prototypical time to pass, and in this brief sample, Golson has been more prone to major errors when placed in this situation.

While I was watching the game footage, it seemed to me that Golson was not throwing as many short passes as he had in the first third of games.  Again, excluding throw aways, the numbers came out as follows:

Length:
Attempts:
Completions:
Comp. Percent.
TD’s
INT:
< 10
26
19
73%
0
0
10+
25
13
52%
1
2

Golson’s completion percentage went up in each subcategory from the same categories in games 1-3.  By the numbers, Golson was basically a 50/50 proposition in terms of whether his throw would travel 10 or more yards through the air or not.  Some of this was self-induced.  Golson displays several pocket tendencies that probably don’t even need to be stated to a Notre Dame fan, but the one in particular I’m thinking of is that amount that he drifts backwards before making a throw.  Sometimes, Golson just overcomplicates things and forces himself to make a much more difficult throw than necessary by expanding the distance between him and the target. 

There was a rather dramatic difference in likelihood of Golson making a throw of 10+ yards between these two groups.  During games 1-3, 66% of his throws were under ten yards in air distance.  Despite 26 fewer qualifying throws, Golson had just one fewer attempts of passes traveling 10+ yards through the air. 

The first chart in this article though indicated that Golson actually averaged fewer yards per completion in games 4-6 than he did in games 1-3.  This seemed odd to me.  He (very) modestly increased his completion percentage in total throws as well as throws of less than and more than 10 yards through the air.  Additionally, he was throwing a greater percentage of his passes a longer distance through the air.  How can this be?  The next chart shows Golson’s average gain per completion for the types of throws mentioned separated by segments:

Throw Length:
Games 1-3:
Games 4-6:
< 10
8.89
7.526
10+
25
16.46

The result is that Golson averaged more yards per completion in each subset during games 1-3 than he did for games 4-6.  This is some function of both actual pass distance as well as yards after catch.  No matter the reason, the truth is that when Golson completed passes in the first three games, Notre Dame was gaining a more significant benefit than what he was able to accomplish in games 4-6. 

Finally, let’s take a look at Golson’s choice in target selection and result:

Target:
Attempts:
Completions:
Comp. Percent.
TD’s:
INT:
WR:
29
17
59%
0
2
RB:
12
9
75%
0
0
TE:
10
6
60%
1
0

Much of Golson’s preference in style of target remained similar.  Wide receiver targets accounted for 56% of Golson’s throws.  He was slightly less successful at hooking up with his receivers (59% compared to 69% in games 1-3), but he increased his completion percentage to both running backs and tight ends.  Teams seemed to continue to buckle down on Eifert as Golson targeted tight ends at a reduced rate (20% compared to 30%), but he was more effective at getting a completion when he targeted his tight ends. 

Golson also improved on his check down and short distance routes to running backs.  Several of Golson’s “pass attempts” to running backs came in the form of chest passes that are similar to toss plays but because the trajectory of the ball was forward from release Golson was given credit for a pass.  If you want to know when this play is unveiled, wait for GA III to take the field because there's a pretty strong correlation.

I’d also thought that Golson would continue to make use of the slip screen/bubble screen plays that he used so well during the first few games.  Boy was I wrong!!  The WR screens were hardly used at all in games 4-6 (something that I think the team would be wise to get back into the normal rotation given its effectiveness).  Golson’s pass distance breakdown to wide receivers follows:

Target Dist.:
Attempts:
Completions:
Comp. Percent.
TD’s:
INT:
+10
18
9
50%
0
2
< 10
11
8
73%
0
0

As opposed to games 1-3 where 70% of his passes intended for receivers travelled less than 10 yards through the air, games 4-6 saw only 38% of his passes to wide receivers travel less than 10 yards.  The designed screen and quick hit plays that Golson thrived on during his first few games were abandoned.  The result was more interceptions and fewer explosive plays.  BH posited that spread offensive schemes like to use these quick hit plays, and I agree.  I’m not entirely sure why these plays were abandoned in games 4-6.  For all of us watching Golson drop back, I think we’d all appreciate a few more designed quick hit plays where the ball gets out of his hands much faster than the “school yard” tactics that frustrate and stress all of us.

I don’t want to make too many observations about this game period outside of what I’ve already said.  Golson had a horrendous game versus Michigan, but if we’re fair, the entire offense had a horrendous game even after Rees came in.  Stanford also frustrated EG’s efforts greatly.  Hopefully progress will exhibit itself when I get to games 7-9 (and yes, I’m aware he didn’t play versus BYU). 

As a means of closing, I’m going to do the same thing I did last article by providing some more general observations, this time with a heavier emphasis on the article subject:

1.  It probably doesn’t take a review of every throw to notice, but EG has a habit of drifting to his right on every play.  We saw BK and Chuck Martin facilitate this tendency in the Pitt game where during OT the ball was consistently placed on the left hash to give him more room to negotiate.  They also do this on kickoffs where a touchback gives them the opportunity to place the ball where the team sees fit.  Golson’s drift limits his ability to utilize the left side of the field, and on more than one occasion eliminated a big play opportunity to the middle or left side of the field.  That said, some of his longest passing gains of the year have directly resulted from him rolling out to the right.

2.  There is a silver-lining to Golson’s scrambling and happy feet aside from the scrambles that sometimes result:  Through the first six games,  I am nearly positive that I haven’t seen a single pass deflection or batted ball at the line of scrimmage. 

3.  The offense began to utilize more designed quarterback runs in games 4-6, and Golson was freed up slightly more to make option plays.  He was not particularly good at it, but that wrinkle was added to the offensive scheme.

4.  Lastly, in yesterday’s article I was pretty critical of Neal’s punt returns in the early going.  I’d like to back off of that a bit.  I’m going to do the (gasp) unthinkable and put a little bit of our punt return troubles on the defense.  Our defensive philosophy of bend but don’t break is not leading to a lot of three and outs.  Our opponents seem to routinely get to at least their 40 yard line or thereabouts before drives stall.  When we add in the routine failures on kick return coverage on poor punting by Ben Turk, I’m more willing to say now that Neal’s not been given a whole lot of opportunities for a clean return.  ND’s opponents are frequently in a position to kick the ball sky high and surround Neal even if we had good returns on.  The few times that Neal’s been given to take a boomer punt back he’s looked considerably more dangerous.

I’ll be back tomorrow with part 3.  Thanks for reading.